This is a competing continuation application to study the effects of parental marital disruption (the first occurrence of separation or divorce) on the short- and long-term well-being of children. The proposed analyses, building on the current project, will focus on divorce not as an event but rather as a process that begins before the parents separate and continues afterward. The great rise in divorce in the 1960s and 1970s means that now, in the early 1990s, the first birth cohorts to experience this high risk of parental divorce are entering adulthood. They will greatly expand the proportion of adults in their 20's and 30' who have experienced a parental divorce. But relatively little is known about the effect of parental divorce during childhood on adult functioning and well-being; and what is known is drawn largely from retrospective reports. One part of this project will be to analyze the 1958 through 1991 waves of the prospective, longitudinal National Child Development Study (NCDS) of Great Britain, which has followed a single birth cohort from age 0 to 33. Key questions will include the following: Do the effects of experiencing a parental divorce-on mental health; on demographic outcomes such as childbearing, union formation, and divorce; and on economic outcomes--persist into mid-life? Does whether or not the parent remarries make a difference? Do late divorces (in the child's late teens or 20's) affect adult children? And do all of these effects still hold once characteristics of the child and his or her family prior to the divorce are taken into account? In addition, assessments were conducted on a random sub-sample of the children of the 33-year-old NCDS panel members in 1991. It is proposed to analyze the effects of parental divorce on this new generation, using the detailed information on their parents from birth to age 33. The key question is: How much of the effect of divorce on children can be accounted for by the history of their parents' psychological functioning since the parents' own childhood? Finally, the effects of divorce on younger children will be pursued, using the 1986-1992 waves of the National Longitudinal Survey of Youth (NLSY). The current project found modest, generally non-significant effects of divorce 2 to 2 1/2 years, on average, after the disruption among children aged 7-16 at the time of the breakup; but it found stronger effects for a heavily-preschool-aged sample 1 year after the breakup. This discrepancy leads to 2 questions that will be explored with the assessments of children in the 1986, 1988, 1990, and 1992 NLSY: Does the disruption itself and its aftermath cause more problems for young (preschool-aged) children than for older children? And/or does the disruption itself and its aftermath produce transitory behavior problems in children of all ages which fade within about two years in most cases? The proposed analyses will begin with simple regression models and hazard models. We then will estimate models that, to the extent possible, take into account sample-selection bias and self-selection into the divorced and non-divorced groups.